Two of the most fundamental mechanisms in biological vision are attention and learning. Attentional mechanisms adaptively select and enhance visual information most relevant to behavior. Neural plasticity allows brain mechanisms to encode visual experience, essential for everyday visual recognition. The challenge is that the amount of visual information to learn is infinite, so attentional mechanisms must select which information to encode. The interplay between attention and visual memory is fundamental for biological vision. Functional magnetic resonance imaging (fMRI) is a powerful tool to study visual attention, learning and memory in the human brain. There is a tight link between stimulus-specific changes in the fMRI signal and the fidelity of visual memory. Prior work has firmly established that attention modulates visual encoding. The current proposal aims to further elucidate the top-down mechanisms that enhance and constrain visual encoding. Aim 1 of this proposal will examine the brain control networks that enhance visual encoding into memory. Memory for items will be selectively enhanced using attention cues. fMRI can then identify the brain regions that show activity differences during encoding of subsequently remembered versus subsequently forgotten items. Aims 2 and 3 will investigate the specificity of visual memory by testing how learning transfers from one instance to another. These aims have implications for understanding how visual training protocols may generalize. In addition, Aim 3 will help to rule out alternative explanations for stimulus-specific changes in the fMRI signal when a visual stimulus is learned, serving to further validate fMRI as a powerful research tool to measure perceptual representations in visual memory. Everyday vision involves recognizing common objects, discriminating friends from strangers, and navigating around familiar environments. The ability to learn new visual information matures throughout development and is active throughout adulthood. Eye disease or damage impairs such basic functions even after sight is restored. Congenital brain disorders or acquired brain damage impair seeing even when the eyes are healthy. This proposal employs fMRI to study how visual learning can be enhanced and what limits its effectiveness. The research has significant implications for theories of visual learning and attention, as well as clinical issues of rehabilitation and recovery from eye disease, injury, or brain damage.